Rational LAMOL: A Rationale-based Lifelong Learning Framework

Kasidis Kanwatchara, Thanapapas Horsuwan, Piyawat Lertvittayakumjorn, Boonserm Kijsirikul, Peerapon Vateekul


Abstract
Lifelong learning (LL) aims to train a neural network on a stream of tasks while retaining knowledge from previous tasks. However, many prior attempts in NLP still suffer from the catastrophic forgetting issue, where the model completely forgets what it just learned in the previous tasks. In this paper, we introduce Rational LAMOL, a novel end-to-end LL framework for language models. In order to alleviate catastrophic forgetting, Rational LAMOL enhances LAMOL, a recent LL model, by applying critical freezing guided by human rationales. When the human rationales are not available, we propose exploiting unsupervised generated rationales as substitutions. In the experiment, we tested Rational LAMOL on permutations of three datasets from the ERASER benchmark. The results show that our proposed framework outperformed vanilla LAMOL on most permutations. Furthermore, unsupervised rationale generation was able to consistently improve the overall LL performance from the baseline without relying on human-annotated rationales.
Anthology ID:
2021.acl-long.229
Volume:
Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)
Month:
August
Year:
2021
Address:
Online
Editors:
Chengqing Zong, Fei Xia, Wenjie Li, Roberto Navigli
Venues:
ACL | IJCNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2942–2953
Language:
URL:
https://aclanthology.org/2021.acl-long.229
DOI:
10.18653/v1/2021.acl-long.229
Bibkey:
Cite (ACL):
Kasidis Kanwatchara, Thanapapas Horsuwan, Piyawat Lertvittayakumjorn, Boonserm Kijsirikul, and Peerapon Vateekul. 2021. Rational LAMOL: A Rationale-based Lifelong Learning Framework. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 2942–2953, Online. Association for Computational Linguistics.
Cite (Informal):
Rational LAMOL: A Rationale-based Lifelong Learning Framework (Kanwatchara et al., ACL-IJCNLP 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.acl-long.229.pdf
Video:
 https://aclanthology.org/2021.acl-long.229.mp4
Code
 kanwatchara-k/r_lamol
Data
BoolQSciFact